Documentos de Académico
Documentos de Profesional
Documentos de Cultura
2, 42-49
Available online at http://pubs.sciepub.com/ajme/4/2/1
Science and Education Publishing
DOI:10.12691/ajme-4-2-1
Abstract The transient radiative heat transfer in two 2D irregular geometry filled with particulate media is
numerically investigated. The radiative transfer equation is solved with FTn finite volume method. Non-orthogonal
mesh is used to discretize the computational domain and the high resolution CLAM scheme is utilized to relate the
facial intensities to the nodal values. Various cases of scattering in media with real indices of refraction (dielectric
particles) and complex indices of refraction (absorbing particles) are considered. Both Mie theory and equivalent
isotropic approximation are used to account for the scattering behavior of the media. The difference between these
two methods is found insignificant, especially for the steady state solutions and for media with complex indices of
refraction, while the complexity and computational cost of the equivalent isotropic approximation is much lower
than those of Mie theory.
Keywords: transient radiative heat transfer, Mie Theory, equivalent isotropic approximation, finite volume method,
non-orthogonal grid
Cite This Article: Elham Khademi Moghadam, Rasool Nasr Isfahani, and Arash Azimi, Numerical
Investigation of the Transient Radiative Heat Transfer inside a Hexagonal Furnace Filled with Particulate
Medium. American Journal of Mechanical Engineering, vol. 4, no. 2 (2016): 42-49. doi: 10.12691/ajme-4-2-1.
2n + 1 z
2
=S1 ( ) n2 + n [ an n (cos ) + bn n (cos )] n (z ) = J n +1 / 2 (z )
n =1 2
(9) (12)
2n + 1 z
2
=S2 ( ) [bn n (cos ) + an n (cos )] n (z ) = H n +1 / 2 (z )
n =1 n
2
+n 2
Where and are related to Legendre polynomials by the On the other hand the Qsca is also can be expressed as
following relations a function of Mie scattering coefficients as follows
dPn (cos )
n (cos ) =
(2n + 1) an 2 + bn 2
2
d (cos ) Qsca = 2
(10) x p n =1
d n (cos )
n (cos ) = sin
2
+ cos n (cos )
d (cos ) (13)
Re(an + bn )
2
Qext = 2
and Mie scattering coefficients an and bn are complex x p n =1
functions of xp=2ap/ and yp=mxp.
If the particles have the same size, the scattering and
n ( y p ) n ( x p ) m n ( y p ) n ( x p )
an = the extinction coefficients of the particles can be
n ( y p ) n ( x p ) m n ( y p ) n ( x p ) expressed as
(11)
m n ( y p ) n ( x p ) n ( y p ) n ( x p ) s, = a 2p NT Qsca
bn = (14)
m n ( y p ) n ( x p ) n ( y p ) n ( x p )
= a 2p NT Qext
Where and are known as Riccati-Bessel functions and
can be related to Bessel and Hankel functions by Although the value of xp is dependent on wavelength,
for the sake of analysis in this study, it is calculated for the
typical wavelength of =3.1415 m as suggested by
Modest [18].
Table 1. media properties for N T=105 (particles/cm3) and =3.1415 m (the values for index of refraction are brought from [8,18])
medium type or particle material xp m (index of refractrion) (m-1) g
F1 5 1.33 1 1 0.8453
F2 2 1.33 1 1 0.6697
B1 1 108 1 1 -0.1794
B2 0.01 108 1 1 -0.4000
Carbon 2 2.20-1.12i 0.979 0.496 0.5990
Anthracite 2 2.05-0.54i 1.013 0.487 0.6549
Bituminous 2 1.85-0.22i 1.051 0.615 0.6456
Lignite 2 1.70-0.066i 0.916 0.822 0.5924
Ash 2 1.50-0.02i 0.574 0.907 0.6341
Carbon 0.5 2.20-1.12i 0.017 0.144 0.0562
Carbon 1 2.20-1.12i 0.224 0.396 0.2146
Carbon 3 2.20-1.12i 1.999 0.507 0.6997
Carbon 5 2.20-1.12i 5.172 0.542 0.7645
Table 2. Dimensionless G (=G/Ebw) along the centerline of the square enclosure (x=0.5 m) predicted by Kim et al. [20]
y(m) isotropic F1 F2 B1 B2
0.020 0.600355 0.518740 0.535868 0.616569 0.632666
0.060 0.556250 0.484921 0.498624 0.570840 0.585263
0.100 0.512708 0.449798 0.461528 0.525645 0.538400
0.140 0.474531 0.419550 0.429650 0.485822 0.496928
0.180 0.441868 0.394810 0.403170 0.451563 0.461070
0.220 0.412683 0.373385 0.379963 0.420857 0.428843
0.260 0.385379 0.353413 0.358333 0.392118 0.398681
0.300 0.359364 0.334146 0.337640 0.364761 0.369998
0.340 0.334617 0.315545 0.317857 0.338761 0.342773
0.380 0.311256 0.297782 0.299127 0.314241 0.317122
0.420 0.289369 0.281022 0.281556 0.291279 0.293117
0.460 0.268964 0.265367 0.265191 0.269885 0.270764
0.500 0.249999 0.250859 0.250021 0.250002 0.249999
0.540 0.232387 0.237473 0.235981 0.231539 0.230725
0.580 0.216013 0.225155 0.222990 0.214376 0.212815
0.620 0.200763 0.213820 0.210952 0.198395 0.196136
0.660 0.186516 0.203371 0.199764 0.183463 0 180560
0.700 0.173155 0.193714 0.189331 0.169461 0.165955
0.740 0.160570 0.184751 0.179558 0.156275 0.152204
0.780 0.148657 0.176397 0.170359 0.143791 0.139191
0.820 0.137300 0.168568 0.161635 0.131900 0.126801
0.860 0.126371 0.161186 0.153281 0.120471 0.114907
0.900 0.115699 0.154161 0.145179 0.109337 0.103343
0.940 0.105067 0.147404 0.137162 0.098291 0.091917
0.980 0.094238 0.140816 0.129025 0.087135 0.080447
45 American Journal of Mechanical Engineering
In addition, the evolution of infinite series in the above The integration of eq. (19) is calculated numerically
equation is terminated by the following relation of using the trapezoid rule with 1000 points in the range of -
Deirmendjian et al. [24]. 11.
n= 9 + 1.2 x p (15)
Table 3. Dimensionless G (=G/Ebw) along the centerline of the square enclosure (x=0.5 m) predicted in this study.
y(m) isotropic F1 F2 B1 B2
0.020 0.599530 0.519407 0.536840 0.614522 0.630658
0.060 0.559019 0.490733 0.503624 0.572443 0.586791
0.100 0.515640 0.456089 0.466700 0.527509 0.540154
0.140 0.473927 0.421347 0.430908 0.484273 0.495285
0.180 0.436937 0.390839 0.399591 0.445811 0.455250
0.220 0.405248 0.365777 0.373455 0.412717 0.420652
0.260 0.377859 0.345202 0.351482 0.384006 0.390520
0.300 0.353372 0.327464 0.332195 0.358285 0.363474
0.340 0.330672 0.311171 0.314405 0.334444 0.338408
0.380 0.309122 0.295505 0.297435 0.311840 0.314682
0.420 0.288479 0.280168 0.281038 0.290226 0.292042
0.460 0.268740 0.265196 0.265231 0.269591 0.270468
0.500 0.250000 0.250780 0.250147 0.250024 0.250039
0.540 0.232367 0.237142 0.235934 0.231623 0.230846
0.580 0.215911 0.224465 0.222704 0.214453 0.212941
0.620 0.200647 0.212856 0.210506 0.198524 0.196327
0.660 0.186538 0.202348 0.199332 0.183790 0.180954
0.700 0.173501 0.192906 0.189123 0.170165 0.166726
0.740 0.161418 0.184443 0.179782 0.157526 0.153516
0.780 0.150144 0.176839 0.171188 0.145724 0.141172
0.820 0.139512 0.169950 0.163200 0.134587 0.129517
0.860 0.129330 0.163622 0.155661 0.123922 0.118357
0.900 0.119384 0.157694 0.148394 0.113519 0.107482
0.940 0.109454 0.152000 0.141198 0.103167 0.096693
0.980 0.099330 0.146370 0.133846 0.092687 0.085831
Then the ability of the code is examined for the case of shown in Figure 2a. All walls are black at Tw=0K. The
non-orthogonal mesh. The transient radiative heat transfer extinction coefficient is 1 m-1. The characteristic time step
in 2D quadrilateral enclosure of Figure 2a is considered. is tc=h/400c where h=1 m for this case. It is found that
The enclosure is filled with an emitting, absorbing and further decrease in tc does not change the results
non-scattering medium at Tref. The 2525 grid system is significantly. The dimensionless radiative heat flux on the
American Journal of Mechanical Engineering 46
bottom wall is compared with previously published results seen. This can secure the accuracy of the results of
of Chai et al. [10]. The comparison is shown in Figure 2b. subsequent sections.
As it can be seen in all cases a good agreement between
the results of present study and previously reported ones is
Figure 1. Scattering phase function for (a) different materials of Table 1 at xp=1 and (b) for Ash at different particle size parameters
Figure 2. The validation check for FT6 FVM with non-orthogonal grid (a) the geometry and grid and (b) the dimensionless heat flux on the bottom wall
radiative intensity reaches at the cold surface. Conversely, approximation and those of Mie theory is 9.0% which
the equivalent optical thickness is lower than the original happens for B2 at t/tc=600. The maximum relative
value for forward scattering. The maximum relative difference for the steady state results (t/tc=800) is 3.6%.
difference between the results of the equivalent isotropic
Figure 3. The temporal dimensionless heat flux on the bottom wall for (a) F1 (the forward scattering medium) and (b) B2 (the backward scattering
medium)
Figure 4. The scattering phase function for (a) different materials of Table 1 at xp=2 and (b) for carbon particles at different size parameters
Then the absorbing particles (those with complex index Figure 5 shows the dimensionless radiative flux on the
of refraction) of Table 1 Are considered. The particle bottom surface. As shown in this figure, the difference in
density is NT=105 particles/cm3. The scattering phase the scattering phase function for various particles is not
function for different materials at xp=2 is shown in Figure 4a. much. Therefore the difference in the values of radiative
The scattering phase function for different sizes of the flux of Figure 5 is mainly attributed to the values of and
carbon particles is also shown in Figure 4b. It is easily . Besides the fact that the radiative intensity travels with
noticed that all particles have the forward scattering lower changes in a media with lower extinction coefficient,
character because their scattering phase functions have the value of single scattering albedo has a significant
higher values for acute angles of scattering. The peak of impact on radiative transfer. In cold regions such as
the function happens at =0. This forward characteristic neighborhood of the bottom wall, the first term on the
becomes more significant for higher size parameters as right of eq. (4) tends to zero and the source term reduces.
can be seen in Figure 4b where the peak of the scattering But the nodal coefficient (aP) remains unchanged.
phase function at =0 becomes more pronounced at Therefore the ratio of s to becomes important and
higher values of xp. higher leads to higher radiative intensity. As a result the
American Journal of Mechanical Engineering 48
heat flux in a medium such as ash with lower value of The heat flux is low at smaller values of x because this
and higher value of must be stronger. region of wall 1 sees more of wall 4 that is a cold wall.
When x/h increases, the influence of wall 4 diminishes
and in the midsection of the bottom wall (around x/h=1),
the heat flux reaches its peak under the influence of hot
wall 3. Then due to the presence of another cold wall (wall
2) the decrement of q starts and at the end of the bottom
wall, it again reaches minimum.
The comparison of Mie calculations with the equivalent
isotropic approximation for media of carbon and lignite
particles is displayed in Figure 6a and Figure 6b,
respectively. Again the heat flux is slightly overpredicted
by the equivalent isotropic approximation. The larger
relative difference between Mie theory and the equivalent
isotropic calculations is 4.3% which occurs for lignite
particles at t/tc=600. The maximum difference has
reduced since in these cases the medium is not purely
scattering and by using the equivalent approximation the
optical thickness does not change as significant as in the
cases of a purely isotropic medium. Also this difference is
larger for medium containing lignite particles because it
has a higher as reported in Table 1. As a result, applying
Figure 5. The temporal dimensionless heat flux on the bottom surface isotropic approximation changes its equivalent and
for different particulate media
more abruptly relative to the media such as carbon
Another noticeable issue in Figure 5 is the influence of particles that have lower value of .
cold walls on the radiative heat flux of the bottom wall.
Figure 6. The temporal dimensionless heat flux on the bottom wall for media of (a) carbon particles and (b) lignite particles
the backward scattering medium. It overpredicts [10] Chai, J.C., Transient radiative transfer in irregular two-
the heat flux for the forward scattering medium; dimensional geometries, J Quant Spectrosc Radiat Transfer, 84,
281-94, 2004.
The radiative flux and integrated intensity are [11] Moghadassian, B., Kowsary, F. and Mosavati, M., "The radiative
higher for media with smaller particles; boundary design of a hexagonal furnace filled with gray and
The equivalent isotropic approximation is much nongray participating gases," Journal of Thermal Science and
simpler than the Mie theory and its predictions Engineering Applications 5(3), 031005, 2013.
[12] Trivic. D.N. and Amon, C.H., Modeling the 3-D radiation of
are accurate especially in the cases of forward anisotropically scattering media by two different numerical
scattering and absorbing particles. Also it is more methods, Int J Heat Mass Transfer, 51, 2711-32, 2008.
accurate for the steady state solutions than the [13] Razi, P., Moghadassian, B. and Shokouhmand, H., "The effect of
transient ones. However, its key parameter, g, is conductive and non-conductive flow diverters on free convection
found when is obtained. Therefore an heat transfer from a vertical array of horizontal isothermal
cylinders," ASME 2009 Heat Transfer Summer Conference
analytical expression or numerical data of is collocated with the InterPACK09 and 3rd Energy Sustainability
required. Conferences. American Society of Mechanical Engineers, 2009.
[14] Moghadassian, B., Razi. P. and Shokouhmand, H., "Investigating
double diffusive convection in an inclined rectangular porous
enclosure subjected to magnetic field," ASME 2009 Heat Transfer
References Summer Conference collocated with the InterPACK09 and 3rd
Energy Sustainability Conferences. American Society of
[1] Chai, J.C., One-dimensional transient radiation heat transfer Mechanical Engineers, 2009.
modeling using a finite-volume method, Numer Heat Transfer B, [15] Steward, F.R. and Trivic, D.N., An assessment of particle
44, 187-208, 2003. radiation in a pulverized-coal-fired boiler, J Inst Energy, 452,
[2] Chai, J.C., Hsu, P-f. and Lam, Y.C., Three-dimensional transient 138-46, 1989.
radiative transfer modeling using the nite-volume method, J [16] Jafari, A., Rahimian M.H. and Saeedmanesh. A., "An unsteady
Quant Spectrosc Radiat Transfer, 86, 299-313, 2004. mixed convection in a driven cavity filled with nanofluids using
[3] Ruan, L.M., Wang, S.G., Qi, H. and Wang, D.L., Analysis of the an externally oscillating lid." Journal of Electronics Cooling and
characteristics of time-resolved signals for transient radiative Thermal Control 3.02, 58, 2013.
transfer in scattering participating media, J Quant Spectrosc [17] Shokouhmand, H., Noori Rahim Abadi S.M.A., and Jafari, A.,
Radiat Transfer, 111, 2405-14, 2010. "The effect of the horizontal vibrations on natural heat transfer
[4] Moghadassian, B., Kowsary, F. and Gholamian, H., "Natural from an isothermal array of cylinders." International Journal of
convection-radiation heat transfer in rectangular cavity with the Mechanics and Materials in Design 7.4, 313-326, 2011.
presence of participating Media," ASME 2010 10th Biennial [18] Modest, M.F., Radiative Heat Transfer. 2nd ed. Academic Press,
Conference on Engineering Systems Design and Analysis. New York, 2003.
American Society of Mechanical Engineers, 2010. [19] Meng, M.P. and Viscanta, R., Radiative transfer in three-
[5] Moghadasian, M. and Moghadassian, B., "3 D inverse boundary dimensional rectangular enclosures containing inhomogeneous
design problem of conduction-radiation heat transfer," Journal of anisotropically scattering media, J Quant Spectrosc Radiat
Applied Sciences(Faisalabad), 12(3), 233-243, 2012. Transfer, 33, 533-49, 1985.
[6] Moghadassian, B. and Kowsary, F., "Inverse boundary design [20] Kim, T.K. and Lee, H., Effect of anisotropic scattering on
problem of natural convectionradiation in a square enclosure," radiative heat transfer in two-dimensional enclosures Int J Heat
International Journal of Thermal Sciences 75, 116-216, 2014. Mass Transfer, 31, 1711-21, 1988.
[7] Kim, M.Y., Menon, S. and Baek, S.W., On the transient radiative [21] Guo, Z. and Kumar, S., Equivalent isotropic scattering
transfer in a one-dimensional planar medium subjected to radiative formulation for transient short-pulse radiative transfer in
equilibrium. Int J Heat Mass Transfer, 53, 5682-91, 2010. anisotropic scattering planar media, Appl Optics, 39, 4411-17,
[8] Trivic, D.N., OBrien, T.J. and Amon, C.H., Modeling the 2000.
radiation of anisotropically scattering media by coupling Mie [22] Van de Hulst, H.C., Light scattering by small particles, Dover
theory with nite volume method, Int J Heat Mass Transfer, 47, publications, New York, 1981.
5765-80, 2004. [23] Kerker, M., The scattering of light and other electromagnetic
[9] Kowsary, F., Gholamian, H. and Moghadassian, B., On the radiation, Academic Press, New York, 1969.
sensitivity of surface reflectance to specularity and phase function [24] Deirmendjian, D., Clasen, R. and Viezee, V., Mie scattering with
parameters in MPA modeling of radiative transfer in two- complex index of refraction, J Opt Soc Am, 51, 620-633, 1961.
component media, Numer Heat Transfer B, 2015.